There's a thread in the modDB forums right now. It's a general survey of what people like, dislike, and wish from of mods.
Interestingly enough, all the members who filled out the survey and posted in the thread seem to have rather similar tastes. There's little difference in the opinions, and as of this posting, no out-right disagreements so far. Maybe this means the modDB forum leans more towards this standpoint than the average modDB user. Either way, some important information came to light.
The most important piece of information was when updating, quality over quantity. Only update when an update is needed. Do not update for the sake of a bump in your mods activity. Specific examples given of a bad update is updating showing some weapon concept art. This ties back into my previous blog post. People don't really care about a new unskinned MP5 or M9, because they've seen a million. This doesn't really show progress in a mod, and in fact it hints to the community that the dev-team lacks creavitity, even if this isn't the case.
This creed of quality over quantity seems important for the entir development process. No-one cares if you get your mod out fast, they want a good mod. All sucessful mods, be they total conversion or minor edits, are done incredibly well. A good developing team takes their time. If a team does this, they'll get loyal fans. If a team overupdates, they'll only get the attention of passer-bys. They'll turn of people who would stay for a long time.
There are a few other things to note. Players don't want a lot of patches; they want 1 or 2 patches that will bring them up to speed.
All in all, quality over quantity and you can't go wrong.